9 Comments
User's avatar
Raja Raman's avatar

Good article Anton. Well articulated. As an EM, It becomes very important to take this up and ensure the team/org don't lose their ways.

Expand full comment
Anton Zaides's avatar

Thank you Raja!

I think the actual 'battle' will happen between CEOs and VP/CTOs, but we as EM can try to show the reality to our own managers.

Expand full comment
Jordan Cutler's avatar

Love this article, Anton. As I'm in developer productivity, I notice a tendency to start goal'ing against the input metrics of AI tool usage and it often worries me. I've been trying to find the right, balance, and this article gives a great resource to point to 🙂

Also big +1 to the overuse of "agent" 😂

Expand full comment
Anton Zaides's avatar

Haha thanks! Not sure what to say in those general agent conversations 😂

Btw your talk in Gdansk is exactly what came to my mind - practical usage in a real world context.

Expand full comment
Jenny Ouyang's avatar

This is so real, Anton, I could barely get through that first section 😅 That’s definitely not the way forward.

It’s kind of hilarious to see some companies enforcing rules of those kinds. Sure, maybe pushing engineers to lean into AI will work out at some point, but I agree with you: AI is still in its early days.

There’s so much more to explore than jumping to black-and-white mandates.

Expand full comment
Anton Zaides's avatar

I think pushing rarely works 😅

As most parents know 😂

Expand full comment
Shubhodaye's avatar

This Article needs to be read by the decision makers who wants to adapt AI without much groundwork.

(The ones who are doing groundwork in implementing AI needs to share their learnings with outside world).

As a tester, testing landscape is going through getting impacted in the way testers test the software with AI.

For me, learning is by doing, sharing what worked, where it failed and how to optimise the next time. Studying the use cases, understanding the biases that AI has been using, testing the data sets etc.,

Keep posting!

Expand full comment
Anton Zaides's avatar

Thank you! 🙏

Expand full comment
YakiUdon's avatar

The hidden assumption here is that generated code is any good and adds value. The first three engineers are human and as gullible as any one else, why would their teammates listen to tales of productivity gains and difficult tasks accomplished simply? We’ve seen broken down bandwagons abandoned roadside many times in this game, and experienced devs will be cautious, if not outright reticent before wasting effort adopting overhyped tools. Turn your team into alpha testers for OpenAI et al? You first, I’ll watch.

The article is like a collection of saccharine coated musings from a willing midwife for corruption. If any ought to be using or be replaced by AI, it is managers like this one.

Expand full comment